Overview of Bedload Estimates Based On Seismic Monitoring At The Ephemeral Arroyo de Los Pinos Tributary of The Rio Grande, New Mexico

The field of environmental seismology encompasses the use of seismic ground motion records produced by important shallow-earth processes, such as glacial motion, subsurface water flow, and bedload transport, to better understand those processes. Seismic monitoring of these processes has several advantages over more traditional surficial process measurements, as seismic data can provide continuous recording over multiple events, potentially at several different positions safely away from the event of interest. Early applications using seismic data to track bedload transport in rivers focused on opportune recordings in the vicinity of high energy perennial rivers; these studies suggest that seismic energy within certain frequency bands is linked to the bedload discharge. Theoretical models of bedload transport in these high energy systems demonstrate the potential for seismic signal frequency characteristics to distinguish bedload transport of different size material. With the relative success of seismic methods in predicting coarse bedload transport in high energy rivers, interest has shifted towards other river systems, such as ephemeral streams that move large quantities of mixed sand-and-gravel sediment during short-lived flash flood events. Even though ephemeral streams in arid regions are common, they are generally data poor because of the infrequency of flood events and their localized occurrence. The Arroyo de los Pinos sediment observatory located in an ephemeral tributary of the Rio Grande is unique in combining a variety of in-channel bedload and water monitoring measurements with dense seismic recording outside of the channel. We have seismic records for over 15 floods ranging from a few centimeters to over one meter of water depth, encompassing bedload flux in the range of 0.2-12 kg s-1 m-1. Our efforts to date have focused on identifying the noise sources within the seismic record, characterization of the seismic properties of the site, and determining the seismic frequency ranges best correlated with the measured bedload flux. Within the frequency range of 30-80 Hz, we find a linear relationship between seismic power and bedload flux. We hypothesize that variations in linear fit statistics between flood events are due to varying grain size distributions in transport and morphological changes in the channel.